Multisensory processing in primary visual cortex: from neural correlates to causation

In our daily conscious lives, we gather information from the external world through the operation of our exteroceptive senses: vision, audition, touch, gustation and olfaction. Our brain processes these streams of information via separate sensory pathways, projecting onto specialized areas of the cerebral cortex such as the primary visual cortex (V1). Yet, the brain also succeeds in combining information from these senses and constructing a unified, conscious percept of our environment and body, a process referred to as multisensory integration. Classically, this integration is thought to occur in higher, associative regions, such as the posterior parietal cortex, or in superior colliculus, a downstream target of primary cortical areas. However, mounting evidence suggests that cross-modal interactions already take place at lower levels, including primary sensory cortices. In mice, behaviorally reported visual stimuli generally elicit a biphasic wave of firing activity in V1, consisting of an early, bottom-up (thalamic-driven) sensory component and a late component depending on top-down (recurrent) feedback. We also found this type of dynamics when presenting multisensory (visual-tactile) stimuli. When we optogenetically assessed the causal requirement of V1 activity in detecting such stimuli, we found that silencing early activity indiscriminately diminished behavioral performance, whereas – surprisingly - silencing late activity suppressed multisensory, but not unisensory (visual-only) detection. This suggests a specific, causal role of late activity in V1 in multisensory perceptual decisions, but we do not yet understand the mechanisms and circuitry by which V1 fulfils this unexpected function. This project aims to unravel: (i) the information coded by V1 neuronal ensembles during late versus early phases of the sensory response; (ii) the cortical brain areas causally involved in generating delayed V1 activity during multisensory processing; (iii) the perceptual conditions under which the causal requirement for delayed V1 activity in multisensory detection arises; (iv) the cell types causally contributing to late V1 activity, and their associated circuitry; (v) the causal importance of these cell types for multisensory perceptual decisions. Using a range of electrophysiological, behavioral, optogenetic and data-analytic techniques in a joint, multi-lab effort, this project aims to contribute to new insights in the functional segregation versus integration of sensory modalities in the cortex. Moreover, this study has the long-term potential to provide clues to develop therapeutic treatments and neuroprosthetics to treat sensory brain damage, for instance after stroke.